Introducing EVMbench
OpenAI and Paradigm release EVMbench, a benchmark for evaluating AI agents on smart contract security tasks including detection, patching, and exploitation of vulnerabilities.
Excerpt
OpenAI and Paradigm introduce EVMbench, a benchmark evaluating AI agents’ ability to detect, patch, and exploit high-severity smart contract vulnerabilities.
Read at source: https://openai.com/index/introducing-evmbench